Introduction

This corpus will encompass the works of Simon and Garfunkel in addition to Paul Simon’s solo work in order to compare the two separate discographies. Paul Simon claims that throughout his time in Simon and Garfunkel the two artists had many creative differences that had to be compromised on however after Simon and Garfunkel parted ways Paul Simon went on to produce many albums of his own written work, now with full creative control. I am curious how these musical differences Paul Simon has alluded to will be measured by Spotify’s AI. Through using a computer to give us actual data about the works of Paul Simon and works of Simon and Garfunkel we can analyze the effect that Art Garfunkel had on the music produced as a band. One obvious difference I predict is a major change in timbre and volume between the two bodies of works as we are comparing two voices harmonizing in Simon and Garfunkel to a solo vocal of Paul Simon. In addition to this I expect timbre and duration will vary as instrumental solos and use of minor or uncommon keys and scales and rhythms were utilized more by Paul Simon in his solo career than as a duo. This being said tracks like “The Boxer”,”50 ways to lose a lover”,and “Somewhere They Cant Find Me” particularly interesting and useful to this analysis as these are fairly popular selections that I feel best represents the artists distinct musical styles. “Bridge Over Troubled Water” is reminiscent of Simon and Garfunkel styles and was a transition period for Paul Simon so comparing this piece of work to that of the duo’s should provide results heavily weighted on lack of harmonies rather than compositional style. In addition to this there will be other differences to account for as I plan for this corpus to be inclusive of as many different recordings of the artists performing their songs, including live recordings which will undoubtedly vary from their studio version matches.


(‘’)

Overview of Track-Level Features


Full Corpus - Energy and Valence Scatterplot

This scatterplot shows a representation of songs by the famous duo Simon and Garfunkel (shown in blue) compared to the solo work of Paul Simon (shown in red) in terms of valence (x-axsis), energy(y-axsis), and loudness(shown through size) as measured by Spotifys AI software. As predicted energy and valence have a positive coorelation regardless of artist similar to energys postive coorelation with loudness depicted on the chart. What I find interesting about this graph is that Paul Simons distribution seems fairly consistent while Simon and Garfunkel have more varation in loudness and energy. In addition I was suprised to see that there is a larger cluster of songs with lower valence in the Simon and Garfunkel songs than in Paul Simons. Generally speaking the highest levels of valence were achieved with energy levels of around 0.60-0.90.

(‘’)

Begining to End - Energy and Valence Scatterplot

This scatterplot is set up the same as the prior however in this one I have singled out the first Simon and Garfunkel album and thwe last Paul Simon album in order to see less clutter and the most contrast. From this plot we can see that there is a larger variation ___.

Clustering a Corpus


Dendrogram

This is a dendrogram attempting to make groups and connectons between the tracks on both Paul Simons latest album and Simon and Garfunkels first album in order to fully understand the contrast between the begining of their duo career and the end of Pauls solo career. Based off of this graph it is clear that ___. This information known, the question now is what makes these groups so different?

General Information; Tempo

***

Tempo Histograms Comparing Artists This is a histogram showing the amount of songs played at certain tempos divided by group. The histogram of Paul Simon shows a generally centered distribution around 120bpm while Simon and Garfunkel is slightly more vaired, being skewed slightly more towards the right with the highest being just around 100bpm. Generally speaking from this histogram we can tell that paul simon has a tendancy to consistently write around 120bpm while the duo generally aimed around 100bpm.

Tempo; Studying Outliers vs Average


Benedictus by Simon and Garfunkel Tempogram

This song was chosen as an outlier for being the lowest in engergy, valence, and sound on the energy and valence scatterplot. The tempogram above shows a tempo of __. __steady or unsteady.

Pigs, Sheep, and Wolves by Paul Simon Tempogram

This song was chosen for its high levels of valence and energy which stands out above others on the energy and valence scatterplots. In the graph it should be noted that the tempo of __ could contribute to its high energy and emotion. __steady or unsteady.

The Sound of Silence by Simon and Garfunkel Tempogram

This song was chosen because it is one of Simon and Garfunkels most popular songs and in the previous energy and valence graph acts as middle ground for the analysis. This tempogram specifically indicates a tempo of __. It also shows that _steady or unsteady.

Generalizations

overall…

Keys


Key Historgrams Comparing Artists These keys histograms show a count of how many times each key was used in the corpus. These show that the two seperate groups are more alike than one might predict. Both groups have a tendancy towards keys G, D, E, and C. That being said G#, F, and A rank much higher with Paul Simon than Simon and Garfunkel. The large increase of the use of G#, A, and E, specifically from Simon and Garfunkels to Paul Simons work implies that these keys were favored by Paul Simon but not Art Garfunkel.

Chromagrams __.

SSM: Timbre and Pitches The Boxer

***

Findings

This page contains a tempogram of the offical recording of the Boxer, by Simon and Garfunkel as well as three different Self Similarity Matrix of different versions of the song, the origonal recording, a live recording by Simon and Garfunkel and a live recording by Paul Simon alone. analysis__.

timbre and loudness

Tab 1: Chromagram


A represenation of the chromatic features of The 59th Bridge Street Song by Simon and Garfunkel.

New Tab 3: Variation Plot


an explanation

New Tab 1.1: Novelty Funct?

Conclusions

This is where my general conclusions will appear once they have been made

Just Messing around

PaulSimonLast <- get_playlist_audio_features("","07cXmheJn34BFNvj9ZL2aT?si=ac471c8aa2584abf")
SimonAndGarfunkelFirst <-get_playlist_audio_features("","249Dhfeva3JzUDeArGoB4y?si=e10d5edc56ce4e98")
FirstandLast <- bind_rows(
  SimonAndGarfunkelFirst %>% mutate(category = "Simon And Garfunkel: Wednesday Morning, 3AM."),
  PaulSimonLast %>% mutate(category = "Paul Simon: In the Blue Light"))

plot_b <-FirstandLast |>
      ggplot(aes(x = valence, y = energy, color = category, size = loudness, text = track.name)) +  geom_point() +   geom_jitter() + labs(x = "Valence", y = "Energy", title = "Energy and Valence Scatterplot")

plot_b <-ggplotly(plot_b)
subplot(plot_b)

OLD Code

PaulSimon<- get_playlist_audio_features("","2WzHYvLkaCRZDjsB04UfSE")
SimonAndGarfunkel<-get_playlist_audio_features("","22geZ3piYiEEYuvRoZ51op")

FullCorpus1<- bind_rows(
     SimonAndGarfunkel |> mutate(category = "Simon And Garfunkel"),
     PaulSimon |> mutate(category = "Paul Simon")
)

FullCorpus1 |>
     ggplot(aes(x = danceability)) +
     geom_histogram(binwidth = 0.1) +
     facet_wrap(~category) + labs(x = "Dance-ability",  y = "Count", title = "Histogram of Dance-ability") 


From this chordogram I would argue that majority of this track is in G major or or E minor. This is a very popular key amongst Paul Simons work. Additionally there is light green line in near the middle, and even brigter near the end where there are not only rests in the guitar and voice parts but also indicating another type of noise, perhaps computorized. This being said I will need to do further analysis when I better grasp the concept… ***

This is a self similarity matrice of Simon and Garfunkels hit, The 59th Bridge Street Song according to timbre. This song was chosen as to represnt a centerpoint. This shows a common theme of fade outs. In this graph it is hard to determine the amount of segments, particuarly towards the middle of the graph although I would say there are 4.


This is a self-similarity matrix on Simon and Garfunkel’s hit The 59th Street Bridge Song based on pitches.